Physical Fitness and Performance

Effects of Tapering on Performance:

A Meta-Analysis
LAURENT BOSQUET1,2, JONATHAN MONTPETIT1, DENIS ARVISAIS1, and IÑIGO MUJIKA3,4
1
Department of Kinesiology, University of Montreal, Montreal, CANADA; 2Faculty of Sport Sciences, University of Lille,
Ronchin, FRANCE; 3Department of Research and Development, Athletic Club Bilbao, Lezama, SPAIN; and 4Department
of Physiology, Faculty of Medicine and Odontology, University of the Basque Country, Álava, SPAIN
Downloaded from http://journals.lww.com/acsm-msse by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 01/24/2021

ABSTRACT
BOSQUET, L., J. MONTPETIT, D. ARVISAIS, and I. MUJIKA. Effects of Tapering on Performance: A Meta-Analysis. Med. Sci.
Sports Exerc., Vol. 39, No. 8, pp. 1358–1365, 2007. Purpose: The purpose of this investigation was to assess the effects of alterations
in taper components on performance in competitive athletes, through a meta-analysis. Methods: Six databases were searched using
relevant terms and strategies. Criteria for study inclusion were that participants must be competitive athletes, a tapering intervention
must be employed providing details about the procedures used to decrease the training load, use of actual competition or field-based
criterion performance, and inclusion of all necessary data to calculate effect sizes. Datasets reported in more than one published study
were only included once in the present analyses. Twenty-seven of 182 potential studies met these criteria and were included in the
analysis. The dependent variable was performance, and the independent variables were the decrease in training intensity, volume, and
frequency, as well as the pattern of the taper and its duration. Pre–post taper standardized mean differences in performance were
calculated and weighted according to the within-group heterogeneity to develop an overall effect. Results: The optimal strategy to
optimize performance is a tapering intervention of 2-wk duration (overall effect = 0.59 T 0.33, P G 0.001), where the training volume is
exponentially decreased by 41–60% (overall effect = 0.72 T 0.36, P G 0.001), without any modification of either training intensity
(overall effect = 0.33 T 0.14, P G 0.001) or frequency (overall effect = 0.35 T 0.17, P G 0.001). Conclusion: A 2-wk taper during which
training volume is exponentially reduced by 41–60% seems to be the most efficient strategy to maximize performance gains. This meta-
analysis provides a framework that can be useful for athletes, coaches, and sport scientists to optimize their tapering strategy.
Key Words: TRAINING INTENSITY, TRAINING VOLUME, TRAINING FREQUENCY, PERIODIZATION
APPLIED SCIENCES

T
he taper is a reduction in the training load of
athletes in the final days before important com-
petition, with the aim of optimizing performance.
This reduction of the training load can be achieved
through the alteration of several components, including
the training volume, intensity, and frequency (72), as well
as the pattern of the taper (i.e., progressive or step taper)
and its duration (25,28,39,46). The taper is widely used by
athletes participating in a wide range of sports differing
in their biomechanical and physiological demands to gain
a performance edge over competitors. In fact, significant
improvements have been reported after tapering for
Address for correspondence: Laurent Bosquet, Ph.D., Department of
Kinesiology, University of Montreal, CP 6128, succ. centre ville, Montreal
(Qc) Canada H3C 3J7; E-mail: laurent.bosquet@montreal.ca.
Submitted for publication June 2006.
Accepted for publication March 2007.
0195-9131/07/3908-1358/0
MEDICINE & SCIENCE IN SPORTS & EXERCISEÒ
Copyright Ó 2007 by the American College of Sports Medicine
DOI: 10.1249/mss.0b013e31806010e0
runners (29), swimmers (48), cyclists (51), rowers (32),
and triathletes (34).
The difficulty for athletes, coaches, and sports scientists
consists in finding the strategy that will maximize the
decrease in accumulated fatigue while retaining or further
enhancing physical fitness, thus leading to peak perfor-
mance. Many strategies to decrease the training load have
been reported in the tapering literature, most of them
leading to an improvement in performance and/or its
physiological correlates (25,28,39,49,55). Some studies
have suggested that the reduction in training volume should
be substantial, somewhere near 85% of normal training
volume (28), whereas others have reported similar improve-
ments in performance after a 31% reduction (3). This
decrease in training volume is generally obtained through
the decrease in the duration of each training session
(3,42,52). However, some studies prefer to manipulate the
training frequency (i.e., the number of training sessions per
week) to decrease weekly training volume (8,31,35). The
duration of taper is also open to wide variations in the
literature. Although most studies have used a 2-wk taper
(3,30,31,34), significant improvements in performance also
have been reported for very short (G7 d) (52) or very long
tapers (928 d) (40).

1358

Copyright @ 2007 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
TABLE 1. Effects of moderator variables on overall effect size for taper-induced The question then arises to determine whether there is an
changes in performance.
optimal strategy suitable for most competitive athletes, as
Overall Effect Size:
Categories Mean (95% CI) N P suggested by Mujika and Padilla (46), or if different patterns
Decrease in training volume of training load reduction can lead to similar improvements
e 20% j0.02 (j0.32, 0.27) 152 0.88 in performance. The purpose of this investigation was to
21–40% 0.27 (0.04, 0.49) 90 0.02
41–60% 0.72 (0.36, 1.09) 118 0.0001
assess the effects of the alterations of taper components on
Q 60% 0.27 (j0.03, 0.57) 118 0.07 performance in competitive athletes, through a systematic
Decrease in training intensity review (or a meta-analysis) of the literature.
Yes j0.02 (j0.37, 0.33) 63 0.91
No 0.33 (0.19, 0.47) 415 0.0001
Decrease in training frequency
Yes 0.24 (j0.03, 0.52) 176 0.08
No 0.35 (0.18, 0.51) 302 0.0001 METHODS
Duration of the taper
e7d 0.17 (j0.05, 0.38) 164 0.14 Literature search. The databases Embase, Kinpubs,
8–14 d 0.59 (0.26, 0.92) 176 0.0005 Physical Education Index, PubMed, SportDiscus, and Web
15–21 d 0.28 (j0.02, 0.59) 84 0.07
Q 22 d 0.31 (j0.14, 0.75) 54 0.18 of Science were searched using the terms (taper* AND
Pattern of the taper (performance* OR competition* OR training) AND (sport*
Step taper 0.42 (j0.11, 0.95) 98 0.12
Progressive taper 0.30 (0.16, 0.45) 380 0.0001
OR exercise* OR swim* OR cycli* OR runn* OR rowi*).
The reference lists of the articles obtained were searched
manually to obtain further studies not identified
In addition to the total decrease in training load, the way to electronically. This led to the identification of 182 potential
decrease it can influence the results, whether it is a single studies for inclusion in the analysis.
stepwise reduction or a progressive exponential reduction with Criteria. Criteria for study inclusion were that
a fast or eventually a slow decay (3). Alternative strategies are participants must be competitive athletes (168 articles
also being tried by athletes in different sports. One of these fulfilled this criterion), the study must employ a tapering
alternative strategies consists in an advanced reduction in the intervention and give all details about the procedures used to
training load, followed by a subsequent increase in the lead- decrease the training load (128 articles), the study must use
up to competition (unpublished observations). The rationale actual competition or field-based criterion performance data
behind this tapering design is that the athlete would take to assess performance capacity (43 articles), and the study
advantage of reduced fatigue levels to enhance training must include all necessary data to calculate effect sizes (i.e.,
tolerance and respond effectively to the training undertaken number of subjects, mean and standard deviation) (32
during the taper, as suggested by recently developed variable articles). Criterion for study exclusion was that the set of data

dose–response mathematical models (6,67). Unfortunately,
no experimental data are available regarding the suitability of
this tapering mode in highly trained athletes.
As indicated by the information above, a wide range of
tapering strategies are currently being used by athletes and
their support teams in view to optimize sports performance.
APPLIED SCIENCES
has been previously published in another article that has
already been included in the present analyses (5 articles). A
total of 27 studies were included in the analysis (2–4,8,14,
16–20,22,26,29–32,34,35,40,42,43,48,51,52,54,63,70).
Coding for the studies. Each study was read and coded
independently by two investigators using the following

TABLE 2. Effects of moderator variables on effect size for taper-induced changes in swimming, running, and cycling performance.
Swimming Running Cycling
Categories Mean (95% CI) N Mean (95% CI) N Mean (95% CI) N
Decrease in training volume
e 20% j0.04 (j0.36, 0.29) 72 No data available 0.03 (j0.62, 0.69) 18
21–40% 0.18 (j0.11, 0.47) 91 0.47 (j0.05, 1.00)‡ 30 0.84 (j0.05, 1.74)‡ 11
41–60% 0.81 (0.42, 1.20)* 70 0.23 (j0.52, 0.98) 14 2.14 (j1.33, 5.62) 15
Q 60% 0.03 (j0.66, 0.73) 16 0.21 (j0.14, 0.56) 66 0.56 (j0.24, 1.35) 36
Decrease in training intensity
Yes 0.08 (j0.34, 0.49) 45 j0.72 (j1.63, 0.19) 10 0.25 (j0.73, 1.24) 8
No 0.28 (0.08, 0.47)* 204 0.37 (0.09, 0.66)* 100 0.68 (0.09, 1.27)† 72
Decrease in training frequency
Yes 0.35 (j0.36, 1.05) 54 0.16 (j0.17, 0.49) 74 0.95 (j0.48, 2.38) 25
No 0.30 (0.10, 0.50)* 195 0.53 (0.05, 1.01)† 36 0.55 (j0.05, 1.15)‡ 55
Duration of the taper
e7d j0.03 (j0.41, 0.35) 54 0.31 (j0.08, 0.70) 52 0.29 (j0.12, 0.70) 47
8–14 d 0.45 (j0.01, 0.90)‡ 84 0.58 (0.12, 1.05)* 38 1.59 (j0.01, 3.19)† 33
15–21 d 0.33 (0.00, 0.65)† 75 j0.08 (j0.95, 0.80) 10 No data available
Q 22 d 0.39 (j0.08, 0.86) 36 j0.72 (j1.63, 0.19) 10 No data available
Pattern of the taper
Step taper 0.10 (j0.65, 0.85) 14 j0.09 (j0.56, 0.38) 36 2.16 (j0.15, 4.47) 25
Progressive taper 0.27 (0.08, 0.45)* 235 0.46 (0.13, 0.80)* 74 0.28 (j0.10, 0.66)‡ 55
* P e 0.01; † P e 0.05; ‡ P e 0.10.

TAPERING OPTIMIZATION Medicine & Science in Sports & Exercised 1359

Copyright @ 2007 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
 moderator variables: the performance capacity, the moda-
lities of the tapering intervention, including the decrease in
training volume, intensity, and frequency, the pattern of the
taper and its duration. An interval scale was used for the
coding of the decrease in training volume and the duration of
the taper, and a nominal scale was used for the remaining
variables. The coding agreement between investigators was
determined by dividing the variables coded the same by the
total number of variables. A mean agreement of 0.90 is
accepted as an appropriate level of reliability in the coding
procedure (58). Mean agreement was 0.927 in our study.
Each coding difference was scrutinized by both investi-
gators together and was resolved before the analysis.
Statistical analysis. Pre–post taper standardized mean
differences in performance were calculated, and weighted
according to the within-group heterogeneity to develop an
overall effect. Statistical significance was set at P G 0.05
level for all analysis. The scale proposed by Cohen (9) was
used for interpretation. The magnitude of the difference was
considered either small (0.2), moderate (0.5), or large (0.8).
All calculations were made with Review Manager 4.2.8
(The Nordic Cochrane Center, The Cochrane Collaboration,
Copenhagen, Denmark).
RESULTS
Overall effect sizes for tapering induced changes in
performance are shown in Table 1 for each moderator
variable. Specific swimming (eight studies, N = 249),
cycling (six studies, N = 80), and running (nine studies,
N = 110) data are presented in Table 2. The dose–response
APPLIED SCIENCES
curves for the overall effect of duration and percentage
decrement in training volume on performance are shown in
Figures 1 and 2, respectively. According to the reported
data, in these competitive athletes, maximal gains are
obtained with a tapering intervention of 2-wk duration,
where the training volume is exponentially decreased by
41–60%, without any modification of either training
intensity or frequency. This generic taper can differ slightly
FIGURE 1—Dose–response curve for the effect of taper duration on
performance.
1360 Official Journal of the American College of Sports Medicine
Copyright @ 2007 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
FIGURE 2—Dose–response curve for the effect of percent decrement
in training volume during taper on performance.
depending on the locomotion mode. The expected perfor-
mance gains are small to moderate for each of these
categories (0.35 G effect size G 0.72, Table 1).
DISCUSSION
The purpose of this investigation was to assess the effects
of the alterations of taper components on performance in
competitive athletes, through a meta-analysis of the liter-
ature. In accordance with previous suggestions (46), we
found that maximal gains are obtained with a tapering
intervention of 2-wk duration, where the training volume is
exponentially decreased by 41–60%, without any modifi-
cation of either training intensity or frequency.
Reduction in training load. Training load can be
decreased through the alteration of training volume,
intensity, and frequency (72). In agreement with previous
suggestions (28,46), this meta-analysis has confirmed that
performance improvement was more sensitive to the
reduction in training volume (Table 1). We computed the
decrease in training volume as the area under the training
volume–time curve. However, some papers did not provide
adequate data to make a precise estimation of this area. It is,
therefore, acknowledged that measurement error is present
at different degrees depending on the studies. We expect
that the use of four categories with relatively wide ranges
will minimize the effect of imprecision on the dose–
response relationship.
Maximal performance gains are obtained with a total
reduction in training volume of 41–60% of pretaper value.
Training volume can be altered through the decrease of the
duration of each training session and/or the decrease of
training frequency. It seems that the first strategy should be
preferred, because decreasing training frequency does not
result in a significant improvement of performance
(Table 1). It should be kept in mind that there is a large
variability between studies, as evidenced by the wide 95%
confidence interval (Table 1). The decrease in training
frequency often interacts with other moderator variables
http://www.acsm-msse.org
like training intensity or the form and duration of the taper,
which makes it difficult to isolate precisely its effect on
performance. A more conservative approach would be to
recommend maintaining training frequency at 80% or more
of the pretaper value (43,46).
As already pointed out by other researchers (28,33,39,
46,55), it seems clear that the training load should not be
reduced at the expense of training intensity (Table 1),
probably because it is a key parameter in the maintenance
of training-induced adaptation during the taper (49). It is
worth noting, however, that the four eligible studies that
decreased training intensity during the taper period were
included in the computation of the overall effect of other
moderator variables. It is legitimate to think that it may
have led to an underestimation of the true overall effect of
these moderator variables. Therefore, we repeated the
analyses by discarding the experimental groups of these
four articles, and we found no difference either on the effect
sizes or on the global dose–response relationship.
Duration of the taper. We found a dose–response
relationship between the duration of the taper and the
performance improvement. Duration of 8–14 d seems to
represent the borderline between the positive influence of
fatigue disappearance and the negative influence of
detraining on performance. Performance improvements
can also be expected after 1-, 3-, or 4-wk tapers (Table 1,
Fig. 1). However, as suggested by 95% confidence
intervals, negative results may be experienced by some
athletes. This interindividual variability in the optimal taper
duration has already been highlighted by some
mathematical modeling studies (40,41,67). Differences in
the physiological and/or psychological adaptation response
to reduced training (40,41,48), as well as the use of an
overload intervention in the weeks before taper (67), are
some of the variables that can account for this variability.
Pattern of the taper. Mujika and Padilla (46) identify
four types of taper patterns: linear taper, exponential taper
with slow or fast time constant of decay of the training load,
and step taper. Because the pattern is not always precisely
detailed in the included studies, we gathered linear and
exponential tapers together into one single pattern named
progressive taper. As can be seen in Table 1, the majority of
studies used a progressive decrease in training load. Our
results agree thoroughly with the results of Banister et al.
(3), who report higher performance improvements after a
progressive taper when compared with a step taper. We
were not able in this study to address the effect of the kind
of progressive taper (i.e., linear or exponential with fast or
slow decay of the training load) on performance. Again,
actual recommendations rely on the work of Banister et al.
(3), who suggest that a fast decay was more beneficial to
performance than a slow decay of the training load.
Expected performance improvements. Tapering is
a training strategy to enhance performance in the most
important athletic competitions. Because performance is a
complex system whose whole is more than the sum of its
TAPERING OPTIMIZATION
Copyright @ 2007 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
parts, specific event performance is the most suitable out-
come to evaluate the effectiveness of a tapering intervention.
This is the reason why all studies using other measures than
actual competition or field-based criterion performance data
to assess performance capacity were not included in the
analyses. When using the scale of Cohen (9) for the inter-
pretation of the effect size, expected performance improve-
ments are most often small, and occasionally moderate.
When expressed as a percent difference, performance
modifications range from j2.28% (54) to 8.91% (19), with
a mean improvement of 1.96%. This difference could be
considered as meaningless if the population of interest were
not competitive athletes. As highlighted by Hopkins et al.
(23), the smallest enhancement of performance that has a
substantial effect on a top athlete’s chance of a medal is
about one third of the typical variation of performance in
competition. This has been shown to be approximately 0.5–
1% in both swimming and running (24,64). In this context,
the gains that can be expected after a taper intervention, as
little as they are, may have a major impact on an athlete_s
success in major competitions. An illustration of this is
provided by Mujika et al. (48), who report that the
magnitude of taper-induced improvements in performance
in the swimming events (2.2%) were of similar order to the
differences between the gold medalist and the fourth place
(1.62%) or between third and eighth place (2.02%) at the
2000 Sydney Olympics.
Possible mechanisms. A number of physiological
changes may account for the tapering-induced improvements
in performance capacity. Maximal oxygen consumption
(V̇O2max) has long been used as a determinant of perfor-
APPLIED SCIENCES
mance (60). The taper-induced hypervolemia and enhanced
red cell production (42,47,62), together with an increase in
oxidative enzyme activity (52,53,62), may contribute
significantly to the increase in oxygen extraction reported
by Neary et al. (54) and, more generally, to the increase in
V̇O2max observed in several studies (16,30,34,51,52).
The energy cost of exercise is another important determi-
nant of performance (15). As reviewed by Mujika et al. (49),
results are disparate, with the discrepancies probably related
to factors such as differences in the training and tapering
programs and the caliber of the athletes. Nevertheless, sig-
nificant improvements have been reported both in swim-
ming (14,31) and running (26,29), but not in cycling. It has
been speculated that biomechanical and neural factors could
explain the taper-induced improvement in the energy cost of
exercise (14,29). The neural hypothesis is consistent with
the increase in strength and power—two factors known to
significantly influence the energy cost of exercise (21), as
has been reported consistently in the taper literature
(22,31,62,68).
The capacity to sustain a high percentage of V̇O2max, also
named aerobic endurance (5), is the last determinant of
performance identified by DiPrampero et al. (15). As
outlined by Péronnet and Thibault (56), the physiological
basis of aerobic endurance is not clearly understood.
Medicine & Science in Sports & Exercised 1361
 Outstanding aerobic endurance can be associated with a
combination of factors, including a high percentage of type
I muscle fibers, the capacity to store large amounts of
muscle and/or liver glycogen, the capacity to spare
carbohydrate by using more fatty acids as energy substrate,
and the capacity to efficiently dissipate heat. If the taper has
been shown to affect significantly the contractile and
metabolic properties of single muscle fibers (52,68), there
are actually no data available to support the hypothesis of a
modification in fiber distribution. It also does not seem that
substrate use is affected, because the respiratory exchange
ratio (RER) is most often unchanged after a taper
intervention (29,51,59). However, the 13–34% taper-
induced increase in muscle glycogen content observed both
in men and women (52,53,62,71) could, undoubtedly,
contribute to improve aerobic endurance.
If these physiological changes can occur very rapidly
when training load is decreased, they also are prone to a rapid
loss when the training stimulus is insufficient (44). In fact,
Coyle et al. (11), as well as Houmard et al. (27), found that
a few days were enough to decrease significantly the blood
volume, which negatively affected the cardiovascular
function by decreasing both the stroke volume (11) and
the cardiac output (13). Enzymatic activity, including
oxidative enzymes and glycogen synthase, decreases also
very rapidly when the training stimulus is insufficient
(12,27,37); this may contribute to the rapid decrease in
V̇O2max and glycogen stores reported in the detraining
literature (10,11,13,27). Altogether, these results show that
taper-induced positive physiological changes are only
transitory and may return very rapidly to pretaper values
APPLIED SCIENCES
or, eventually, to initial values in the case of a long-term
insufficient training stimulus (45). Whence the difficulty to
find the taper strategy that will allow the athlete to recover
and overcompensate adequately from prior heavy training
loads while avoiding detraining.
Practical considerations. According to our results, a
tapering intervention of 2-wk duration, where the training
volume is exponentially decreased by 41–60%, without any
modification of either training intensity or frequency, is the
strategy that will maximize the probability to obtain sig-
nificant improvements in performance. From a practical
point of view, it is important to determine whether this
‘‘optimal’’ taper is a generic one, or whether it differs accord-
ing to the sport, the gender, or the fatigue status before
the taper.
Specific swimming, cycling, and running data are shown
in Table 2. Rowing was not considered because the total
number of subjects (two studies, N = 23) was insufficient to
address all the moderator variables used in this meta-
analysis. Although the number of swimming, cycling or
running subjects per moderator variable was also insuffi-
cient to provide each sport with specific recommendations,
it was possible to identify some trends. The first, indis-
putable one is the need to maintain training intensity,
whatever the mode of locomotion. If a 41–60% decrease in
1362 Official Journal of the American College of Sports Medicine
Copyright @ 2007 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
training volume seems to be optimal in swimming, we were
not able to find a similar cutoff value in cycling and run-
ning, the optimal decrease being somewhere between 21
and 60%. A period of 8–14 d seems to represent the optimal
taper duration in cycling and running. It should be noted,
however, that significant improvements can be expected
from longer taper durations in swimming, but also that the
number of cycling and running subjects for such durations
is insufficient to test this hypothesis with adequate statistical
power (N = 10 in running and 0 in cycling for 15–21 d and
22 d or more). Finally, there seems to be some controversy
in cycling about the pattern of taper that should be pre-
ferred, but also to determine whether a decrease in training
frequency should be used to decrease training volume.
Because sex has been shown to affect some of the
mechanisms possibly leading to an improvement in perform-
ance, such as the capacity to increase glycogen stores (66), it
is quite legitimate to question the applicability of this stra-
tegy to women. Among the 27 included studies, only nine
of them used a mixed sample, and only three of them pro-
vided separated results that would allow comparisons be-
tween males and females. Smith (63) did not report any sex
difference in the amplitude of performance improvement
after a 1-wk taper in 6 female and 10 male elite rowers.
These results were in accordance with Mujika et al. (40),
who found no difference between 8 female and 10 male
elite swimmers after two tapers of 3- and 4-wk duration. In
a more recent study, Mujika et al. (48) registered the per-
formance change of 49 female and 50 male elite swimmers
in the final 3-wk taper leading to the Sydney 2000 Olympic
Games. They found a higher improvement in males than in
females in freestyle events (2.54 T 1.5 vs 1.60 T 1.33%, P G
0.05), but not in form events (backstroke, breaststroke,
butterfly). The observational nature of this study did not
allow the authors to provide any psychological and phy-
siological explanation for this difference. Other confound-
ing factors such as shaving or the use of different swimming
suits may have affected the results. With the exception of
this possible sex effect on the magnitude of taper-induced
performance improvement, we found no apparent evidence
questioning the generalization of our results to women.
Another important practical consideration is whether the
optimal taper depends on the severity of fatigue an athlete
carries into the taper process. The theoretical modeling work
of Thomas and Busso (67) clearly shows that using an over-
load procedure before the taper results in higher performance
gains, but also that taper duration and percentage decrease in
training load should be adapted (i.e., increased) to dissipate
this extra accumulated fatigue. Despite this sound theoretical
background, very few experimental studies addressed this
topic. This situation probably arises from the difficulty to
quantify fatigue. The overreaching/overtraining literature of
the past two decades clearly indicates that there exists no
pathognomonic clinical sign of severe fatigue (69). An
affordable and very interesting tool to assess fatigue levels
is probably the profile of mood states (POMS) (36). If
http://www.acsm-msse.org
several taper-studies included the POMS in their measures
(4,17,22,34,38,57), only two of them used it as a tool to
quantify initial fatigue (4,34). Margaritis et al. (34) used a
total mood disturbance (TMD) score with the purpose of
identifying overreaching or overtraining. However, they did
not provide cutoff values that would help to classify the
subjects into one of these categories. In the study by Berger
et al. (4), eight elite pursuit cyclists completed the POMS and
a simulated 4-km pursuit performance test throughout a 6-wk
period, including a 1-wk baseline, a 3-wk overload, and a 2-
wk taper. They found the TMD score to increase after the
overload period and to decrease below baseline values after
the taper period. If it is clear from this study and many others
(4,17,22,34,38,57) that TMD or subscales scores are sensi-
tive to fatigue, there is still difficulty in delineating cutoff
values that would allow classification of subjects or
comparison of studies.
Athletes_ diet also may affect the benefits that can be
expected from a well-designed taper. Margaritis et al. (34)
have reported the daily energy intake, energy expenditure,
body mass, and body fat of 20 male long-distance triathletes
during 4 wk of overloaded training and then 2 wk of taper.
Energy intake did not change between both phases (15.0 T
3.2 vs 15.0 T 3.3 MJIdj1), whereas energy expenditure
decreased from 17.0 T 1.9 to 12.1 T 1.2 MJIdj1 (P G 0.05).
Total body mass did not change during the taper (67.9 T 3.6
vs 67.8 T 3.2 kg, NS), but percent body fat increased from
11.4 T 3.7 to 11.8 T 3.7% (P G 0.05). Consequently, care
should be taken to match energy intake with the reduced
energy expenditure that characterizes the taper. Muscle
glycogen concentration has been shown to increase during
the taper (52,53,71). Because the amplitude of glycogen
supercompensation has been associated with the amplitude
of performance improvement in a 40-km time trial (52), it is
important to use nutritional strategies that will maximize the
replenishment of glycogen stores. Walker et al. (71) found
in six eumenorrheic women that glycogen supercompensa-
tion and performance in a constant-power test at 80%
V̇O2max were higher after a 7-d taper with a high-
carbohydrate diet (78% carbohydrate) during the final 4 d
compared with a moderate-carbohydrate diet (48% carbo-
hydrate) for 7 d. Consequently, a rich carbohydrate diet
seems to be an important component of a successful taper.
Limits. Specific event competition or field-based
criterion performance data were designed as the dependant
variable to test the effectiveness of taper strategies. Correct
interpretation of our results relies on the assumption that
both pre- and posttaper performances were indeed maximal.
Although this methodological consideration was of primary
importance, we found very few studies using a criterion to
control this parasite variable. Neary et al. (51,52,54) made
an effort in this direction by taking the rating of perceived
exertion at the end of each performance test. They reported
no difference between pre- and posttaper values, suggesting
that efforts were at least of similar difficulty. Houmard et al.
(29) computed the intraclass coefficient of correlation (ICC)
TAPERING OPTIMIZATION
Copyright @ 2007 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
between pre- and posttaper values to assess the consistency
between datasets. They reported an ICC of 0.99, suggesting
that both absolute and relative errors were low. The fact that
other studies included in this meta-analysis did not control
this point probably limits the interpretation that can be made
from their results. In fact, a possible placebo effect can
interact with a natural tendency to go deeper during the last
performance test and lead to an overestimation of taper-
induced performance gains. It should be noted, however, that
10 of the 27 studies included in this meta-analysis used
competition data (2,17,18,20,22,31,40,42,43,48), where
motivation to give an all-out effort was probably the same
for each test because, in addition to the competitive stake,
they were part of a more global competition schedule. More-
over, the 17 remaining studies used field-based performance
tests that have been shown to be very reliable (1,50,61,65),
thus limiting the error of measurement.
The aim of a meta-analysis is to combine the body of
literature on a given topic to propose evidence-based
conclusions. In this study we identified numerous moderator
variables to describe as close as possible the diversity of
tapering strategies currently used by coaches and sport
scientists. A prerequisite to make valid comparisons is to
have roughly the same number of subjects per moderator
variables. With the exception of the decrease in training
volume, most of the remaining independent variables did not
meet this condition. Although a duration of 8–14 d seems to
be optimal in terms of performance gains, it has to be
recognized that additional data are needed to test the real
effectiveness of longer taper durations (N = 164, 176, 84,
and 54 for 1-, 2-, 3-, or 4-wk taper durations, respectively).
APPLIED SCIENCES
The choice of a given pattern is also complicated by the
difference in the number of subjects (N = 98 and 380 for
step and progressive taper, respectively). A lower statistical
power make step taper performance gains nonsignificant,
whereas the overall effect is higher than that of progressive
taper (Table 1). The same is true for the decrease in training
frequency. Future studies should test these conditions, to
verify whether the predictions of this meta-analysis are true
even when the number of subjects per moderator variable is
not always adequate.
The inclusion of studies using exclusively competitive
athletes probably introduced a bias in our meta-analysis. In a
context where no one wants to take the chance of a sub-
standard performance, it is likely that researchers are study-
ing tapering strategies that have been proved to be successful,
either scientifically or empirically, thus leading to a circular
thinking. Innovative tapering designs should be tested to
have a comprehensive understanding of the adaptive re-
sponse to the decrease in training load. New developments in
mathematical modeling offer an interesting alternative
(6,7,67). As underscored by Thomas and Busso (67), mathe-
matical simulations provide a convenient technique to deter-
mine the combination of reduction in training load and its
duration that will maximize performance, simply by chang-
ing the pattern of training. This objective tool could be used
Medicine & Science in Sports & Exercised 1363
 to propose potentially successful innovative tapering designs
that will be tested with competitive athletes afterwards.
CONCLUSION
The purpose of this investigation was to assess the effects
of the alterations of taper components on performance in
competitive athletes, through a meta-analysis of the liter-
ature. A 2-wk taper during which training volume is
exponentially reduced by 41–60% without altering training
intensity or frequency appears to be the most efficient
REFERENCES
1. ALBERTY, M., M. SIDNEY, F. HUOT-MARCHAND, et al. Reproducibility
of performance in three types of training test in swimming. Int. J.
Sports Med. 27:623–628, 2006.
2. ATLAOUI, D., M. DUCLOS, C. GOUARNE, L. LACOSTE, F. BARALE, and
J. C. CHATARD. 24-hr urinary catecholamine excretion, training and
performance in elite swimmers. Int. J. Sports Med. 27:314–321,
2006.
3. BANISTER, E. W., J. B. CARTER, and P. C. ZARKADAS. Training
theory and taper: validation in triathlon athletes. Eur. J. Appl.
Physiol. 79:182–191, 1999.
4. BERGER, B., R. MOTL, B. BUTKI, D. MARTIN, J. WILKINSON, and
D. OWEN. Mood and cycling performance in response to three
weeks of high-intensity, short-duration overtraining, and a two-
week taper. Sport Psychol. 13:444–457, 1999.
5. BOSQUET, L., L. LEGER, and P. LEGROS. Methods to determine
aerobic endurance. Sports Med. 32:675–700, 2002.
6. BUSSO, T. Variable dose-response relationship between exercise
training and performance. Med. Sci. Sports Exerc. 35:1188–1195,
2003.
APPLIED SCIENCES
7. BUSSO, T., and R. THOMAS. Using mathematical modeling in
training planning. Int. J. Sports Physiol. Perf. 1:400–405, 2006.
8. CHILD, R. B., D. M. WILKINSON, and J. L. FALLOWFIELD. Effects of a
training taper on tissue damage indices, serum antioxidant
capacity and half-marathon running performance. Int. J. Sports
Med. 21:325–331, 2000.
9. COHEN, J. Statistical Power Analysis for the Behavioral Sciences.
Hillsdale, NJ: Lawrence Erlbaum Associates, 1988.
10. COSTILL, D., D. KING, and R. THOMAS. Effects of reduced training
on muscular power in swimmers. Phys. Sports Med. 13:94–101,
1985.
11. COYLE, E. F., M. K. HEMMERT, and A. R. COGGAN. Effects of
detraining on cardiovascular responses to exercise: role of blood
volume. J. Appl. Physiol. 60:95–99, 1986.
12. COYLE, E. F., W. H. MARTIN 3RD, S. A. BLOOMFIELD, O. H. LOWRY,
and J. O. HOLLOSZY. Effects of detraining on responses to
submaximal exercise. J. Appl. Physiol. 59:853–859, 1985.
13. COYLE, E. F., W. H. MARTIN 3RD, D. R. SINACORE, M. J. JOYNER,
J. M. HAGBERG, and J. O. HOLLOSZY. Time course of loss of
adaptations after stopping prolonged intense endurance training.
J. Appl. Physiol. 57:1857–1864, 1984.
14. D_ACQUISTO, L. J., M. BONE, S. TAKAHASHI, G. LANGHANS, A. P.
BARZDUKAS, and J. P. TROUP. Changes in aerobic power and
swimming economy as a result of reduced training volume. Swim
Sci. VI:201–205, 1992.
15. DI PRAMPERO, P. E., G. ATCHOU, J. C. BRUCKNER, and C. MOIA. The
energetics of endurance running. Eur. J. Appl. Physiol. Occup.
Physiol. 55:259–266, 1986.
16. DRESSENDORFER, R. H., S. R. PETERSEN, S. E. MOSS LOVSHIN, J. L.
HANNON, S. F. LEE, and G. J. BELL. Performance enhancement with
1364 Official Journal of the American College of Sports Medicine
Copyright @ 2007 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
strategy to maximize performance gains. Alternative tapering
designs could also have a beneficial effect on performance of
individual athletes, but more pronounced interindividual
differences may be expected. This meta-analysis provides
a framework that can be useful for athletes, coaches and
sport scientists to optimize their tapering strategy. Future
investigations should evaluate alternative and innovative
tapering strategies exclusive to those included in this meta-
analysis, which could prove to provide further performance
benefits to athletes, as suggested by recent mathematical
modeling simulations.
maintenance of resting immune status after intensified cycle
training. Clin. J. Sport Med. 12:301–307, 2002.
17. FLYNN, M. G., F. X. PIZZA, J. B. BOONE JR., F. F. ANDRES, T. A.
MICHAUD, and J. R. RODRIGUEZ-ZAYAS. Indices of training stress
during competitive running and swimming seasons. Int. J. Sports
Med. 15:21–26, 1994.
18. GARET, M., N. TOURNAIRE, F. ROCHE, et al. Individual interdepen-
dence between nocturnal ANS activity and performance in
swimmers. Med. Sci. Sports Exerc. 36:2112–2118, 2004.
19. HALSON, S. L., M. W. BRIDGE, R. MEEUSEN, et al. Time course
of performance changes and fatigue markers during intensi-
fied training in trained cyclists. J. Appl. Physiol. 93:947–956,
2002.
20. HARBER, M. P., P. M. GALLAGHER, A. R. CREER, K. M. MINCHEV,
and S. W. TRAPPE. Single muscle fiber contractile properties
during a competitive season in male runners. Am. J. Physiol.
287:R1124–R1131, 2004.
21. HOFF, J., J. HELGERUD, and U. WISLOFF. Maximal strength traiing
improves work economy in trained female cross-country skiers.
Med. Sci. Sports Exerc. 31:870–877, 1999.
22. HOOPER, S. L., L. T. MACKINNON, and E. M. GINN. Effects of three
tapering techniques on the performance, forces and psychometric
measures of competitive swimmers. Eur. J. Appl. Physiol. 78:
258–263, 1998.
23. HOPKINS, W. G., J. A. HAWLEY, and L. M. BURKE. Design and
analysis of research on sport performance enhancement. Med. Sci.
Sports Exerc. 31:472–485, 1999.
24. HOPKINS, W. G., and D. J. HEWSON. Variability of competitive
performance of distance runners. Med. Sci. Sports Exerc.
33:1588–1592, 2001.
25. HOUMARD, J. A. Impact of reduced training on performance in
endurance athletes. Sports Med. 12:380–393, 1991.
26. HOUMARD, J. A., D. L. COSTILL, J. B. MITCHELL, S. H. PARK, R. C.
HICKNER, and J. N. ROEMMICH. Reduced training maintains
performance in distance runners. Int. J. Sports Med. 11:46–52,
1990.
27. HOUMARD, J. A., T. HORTOBAGYI, R. A. JOHNS, et al. Effect of short-
term training cessation on performance measures in distance
runners. Int. J. Sports Med. 13:572–576, 1992.
28. HOUMARD, J. A., and R. A. JOHNS. Effects of taper on swim
performance. Practical implications. Sports Med. 17:224–232,
1994.
29. HOUMARD, J. A., B. K. SCOTT, C. L. JUSTICE, and T. C. CHENIER.
The effects of taper on performance in distance runners. Med. Sci.
Sports Exerc. 26:624–631, 1994.
30. JEUKENDRUP, A. E., M. K. HESSELINK, A. C. SNYDER, H. KUIPERS,
and H. A. KEIZER. Physiological changes in male competitive
cyclists after two weeks of intensified training. Int. J. Sports Med.
13:534–541, 1992.
http://www.acsm-msse.org
31. JOHNS, R. A., J. A. HOUMARD, R. W. KOBE, et al. Effects of taper on
swim power, stroke distance, and performance. Med. Sci. Sports
Exerc. 24:1141–1146, 1992.
32. JURIMAE, J., J. MAESTU, and T. JURIMAE. Leptin as a marker of
training stress in highly trained male rowers? Eur. J. Appl.
Physiol. 90:533–538, 2003.
33. KUBUKELI, Z. N., T. D. NOAKES, and S. C. DENNIS. Training
techniques to improve endurance exercise performances. Sports
Med. 32:489–509, 2002.
34. MARGARITIS, I., S. PALAZZETTI, A. S. ROUSSEAU, M. J. RICHARD, and
A. FAVIER. Antioxidant supplementation and tapering exercise
improve exercise-induced antioxidant response. J. Am. Coll. Nutr.
22:147–156, 2003.
35. MCCONELL, G. K., D. L. COSTILL, J. J. WIDRICK, M. S. HICKEY,
H. TANAKA, and P. B. GASTIN. Reduced training volume and
intensity maintain aerobic capacity but not performance in
distance runners. Int. J. Sports Med. 14:33–37, 1993.
36. MCNAIR, D., M. LORR, and L. DROPPLEMAN. Profile of Mood States
Manual. San Diego, CA: Educational and Industrial Testing
Services, 1971.
37. MIKINES, K. J., B. SONNE, B. TRONIER, and H. GALBO. Effects of
acute exercise and detraining on insulin action in trained men.
J. Appl. Physiol. 66:704–711, 1989.
38. MORGAN, W. P., D. R. BROWN, J. S. RAGLIN, P. J. O_CONNOR, and
K. A. ELLICKSON. Psychological monitoring of overtraining and
staleness. Br. J. Sports Med. 21:107–114, 1987.
39. MUJIKA, I. Influence of training characteristics and tapering on the
adaptation in highly trained individuals: a review. Int. J. Sports
Med. 19:439–446, 1998.
40. MUJIKA, I., T. BUSSO, L. LACOSTE, F. BARALE, A. GEYSSANT, and
J. CHATARD. Modeled responses to training and taper in compet-
itive swimmers. Med. Sci. Sports Exerc. 28:251–258, 1996.
41. MUJIKA, I., J. CHATARD, T. BUSSO, A. GEYSSANT, F. BARALE, and
L. LACOSTE. Use of swim-training profiles and performances data
to enhance training effectiveness. J. Swim Res. 11:23–29, 1996.
42. MUJIKA, I., A. GOYA, S. PADILLA, A. GRIJALBA, E. GOROSTIAGA, and
J. IBANEZ. Physiological responses to a 6-d taper in middle-
distance runners: influence of training intensity and volume. Med.
Sci. Sports Exerc. 32:511–517, 2000.
43. MUJIKA, I., A. GOYA, E. RUIZ, A. GRIJALBA, J. SANTISTEBAN, and
S. PADILLA. Physiological and performance responses to a 6-day
taper in middle-distance runners: influence of training frequency.
Int. J. Sports Med. 23:367–373, 2002.
44. MUJIKA, I., and S. PADILLA. Detraining: loss of training-induced
physiological and performance adaptations. Part I: short term
insufficient training stimulus. Sports Med. 30:79–87, 2000.
45. MUJIKA, I., and S. PADILLA. Detraining: loss of training-induced
physiological and performance adaptations. Part II: long term
insufficient training stimulus. Sports Med. 30:145–154, 2000.
46. MUJIKA, I., and S. PADILLA. Scientific bases for precompetition
tapering strategies. Med. Sci. Sports Exerc. 35:1182–1187, 2003.
47. MUJIKA, I., S. PADILLA, and A. GEYSSANT. Hematological responses
to training and taper in competitive swimmers: relationships with
performance. Arch. Physiol. Biochem. 105:379–385, 1997.
48. MUJIKA, I., S. PADILLA, and D. PYNE. Swimming performance
changes during the final 3 weeks of training leading to the Sydney
2002 Olympic Games. Int. J. Sports Med. 23:582–587, 2002.
49. MUJIKA, I., S. PADILLA, D. PYNE, and T. BUSSO. Physiological
changes associated with the pre-event taper in athletes. Sports
Med. 34:891–927, 2004.
50. NEARY, J. P., G. J. BELL, and H. A. QUINNEY. Reproducibility in
simulated 40 km time trial and its associated metabolic stress.
Sports Med. Train. Rehabil. 9:79–88, 1999.
51. NEARY, J. P., Y. N. BHAMBHANI, and D. C. MCKENZIE. Effects of
different stepwise reduction taper protocols on cycling perfor-
mance. Can. J. Appl. Physiol. 28:576–587, 2003.
TAPERING OPTIMIZATION
Copyright @ 2007 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.
52. NEARY, J. P., T. P. MARTIN, and H. A. QUINNEY. Effects of taper on
endurance cycling capacity and single muscle fiber properties.
Med. Sci. Sports Exerc. 35:1875–1881, 2003.
53. NEARY, J. P., T. P. MARTIN, D. C. REID, R. BURNHAM, and H. A.
QUINNEY. The effects of a reduced exercise duration taper
programme on performance and muscle enzymes of endurance
cyclists. Eur. J. Appl. Physiol. Occup. Physiol. 65:30–36, 1992.
54. NEARY, J. P., D. C. MCKENZIE, and Y. N. BHAMBHANI. Muscle
oxygenation trends after tapering in trained cyclists. Dyn. Med.
4:4, 2005.
55. NEUFER, P. D. The effect of detraining and reduced training on the
physiological adaptations to aerobic exercise training. Sports Med.
8:302–321, 1989.
56. PÉRONNET, F., and G. THIBAULT. Mathematical analysis of running
performance and world running records. J. Appl. Physiol. 67:
453–465, 1989.
57. RAGLIN, J. S., W. P. MORGAN, and P. J. O_CONNOR. Changes in
mood states during training in female and male college swimmers.
Int. J. Sports Med. 12:585–589, 1991.
58. RHEA, M. R., B. A. ALVAR, L. N. BURKETT, and S. D. BALL. A
meta-analysis to determine the dose response for strength
development. Med. Sci. Sports Exerc. 35:456–464, 2003.
59. RIETJENS, G. J., H. A. KEIZER, H. KUIPERS, and W. H. SARIS. A
reduction in training volume and intensity for 21 days does not
impair performance in cyclists. Br. J. Sports Med. 35:431–434,
2001.
60. SALTIN, B., and P. O. ASTRAND. Maximal oxygen uptake in
athletes. J. Appl. Physiol. 23:353–358, 1967.
61. SCHABORT, E. J., J. A. HAWLEY, W. G. HOPKINS, and H. BLUM. High
reliability of performance of well-trained rowers on a rowing
ergometer. J. Sports Sci. 17:627–632, 1999.
62. SHEPLEY, B., J. D. MACDOUGALL, N. CIPRIANO, J. R. SUTTON, M. A.
TARNOPOLSKY, and G. COATES. Physiological effects of tapering in
highly trained athletes. J. Appl. Physiol. 72:706–711, 1992.
63. SMITH, H. K. Ergometer sprint performance and recovery with
variations in training load in elite rowers. Int. J. Sports Med.
21:573–578, 2000.
APPLIED SCIENCES
64. STEWART, A. M., and W. G. HOPKINS. Consistency of swimming
performance within and between competitions. Med. Sci. Sports
Exerc. 32:997–1001, 2000.
65. STICKLAND, M., S. PETERSEN, and R. DRESSENDORFER. Critical
aerobic power during simulated 20 km bicycle racing. Sports
Med. Train. Rehabil. 9:289–301, 2000.
66. TARNOPOLSKY, M. A., S. A. ATKINSON, S. M. PHILLIPS, and J. D.
MACDOUGALL. Carbohydrate loading and metabolism during
exercise in men and women. J. Appl. Physiol. 78:1360–1368,
1995.
67. THOMAS, L., and T. BUSSO. A theoretical study of taper character-
istics to optimize performance. Med. Sci. Sports Exerc. 37:
1615–1621, 2005.
68. TRAPPE, S., D. COSTILL, and R. THOMAS. Effect of swim taper on
whole muscle and single muscle fiber contractile properties. Med.
Sci. Sports Exerc. 32:48–56, 2000.
69. URHAUSEN, A., and W. KINDERMANN. Diagnosis of overtraining:
what tools do we have? Sports Med. 32:95–102, 2002.
70. VAN HANDEL, P. J., A. KATZ, J. TROUP, J. DANIELS, and P. BRADLEY.
Oxygen consumption and blood lactic acid response to training
and taper. In: Swimming Science V. Champaign, IL: Human
Kinetics, pp. 269–275, 1988.
71. WALKER, J. L., G. J. HEIGENHAUSER, E. HULTMAN, and L. L. SPRIET.
Dietary carbohydrate, muscle glycogen content, and endurance
performance in well-trained women. J. Appl. Physiol. 88:
2151–2158, 2000.
72. WENGER, H. A., and G. J. BELL. The interactions of intensity,
frequency and duration of exercise training in altering cardiores-
piratory fitness. Sports Med. 3:346–356, 1986.
Medicine & Science in Sports & Exercised 1365